have falltn from the HeaveHs, 351 



The method employed to feparate the iron and the mag- 

 nefia is founded on this circumftance, that a folution of that 

 earth containing an excefs of acid is not precipitated by am- 

 monia, becaufe there is formed a triple faU, which cannot be 

 decompofed by a fuperabundance of that alkali ; but the acid 

 muft be ia fuch quantity that the fait refulting from its com- 

 bination with the ammonia may be able to faturate the fait 

 of magnefia exifting in the liquor. Care was taken to obferve 

 this. 



Alkaline carbonates produce no precipitation of the fub- 

 ftances which enter into combination with the triple fait here 

 mentioned, though it muft, however, be decompofed ; but 

 there is formed another kind, compofed of magnefia, car- 

 bonic acid, and ammonia, which remains in folution. But 

 one thing I did not forefee was the fimultaneous precipita- 

 tion of a portion of nickel and magnefia by the cauftic pot- 

 a(h ; for, as Mr, Howard remarked, this metal is found there 

 in complex combination, the oxide of which ought to be 

 diflblved by the ammonia in proportion as it becomes free : 

 the precipitation of this metal muft be produced by its affi- 

 nity for magnefia, at leaft this is what appears moft probable. 

 After theie two analyfes of the ftone of Benares, and the 

 conformity of the refults which they furnifhed, though ef- 

 fected by methods fomewhat different, there can be no doubt 

 that it is really compofed of iron, filex, magnefia, nickel, and 

 fulphur, as Mr. Howard announced. 



Having therefore confirmed by thefe two means, and by 

 others, which it is necdlefs to mention, the refults given by 

 the Englifti chemift in regard to the compofition of the ftone 

 of Benares, I fubje6led to fimilar experiments two other kinds 

 of ftone which fell in France, in order to afcertain whether 

 they would furnifti the fame principles of analyfis as their 

 external characters feemed to announce. 



To avoid a repetition of the details already given of the 

 analytical methods employed, I fliall only obferve, that the 

 ftones which fell at Barbotan and Juliac, treated with ful- 

 phuric, nitric, muriatic, Sec. acids, exhibited in the feries of 

 operations to which they were fubjedted, and by the nature 

 of the elements wliich they furnidied, the moft perfcdl fiini- 

 larity to the preceding. It may therefore be confidcred as 

 fully proved, that the ftones faid to have fallen from the 

 clouds in ditfcrent countries are compofed of principle.-; pcr- 

 h&.\y fimilar, and confequently that we muft have rccourle 

 to a common caufe to explain their formation, and to fup.- 

 pofe alfo a common fource from which nature derived the 

 «lemcnts of them. I muft, however, acknowledge, ihit the 



Ipeciaiens 



